INDEX

Explanations

keywords related to control and authority in various contexts

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

gow

-0.72

istle

-0.68

 Honour

-0.66

ruary

-0.66

neys

-0.66

Recommend

-0.65

à¤

-0.65

deals

-0.64

Stand

-0.64

hire

-0.64

POSITIVE LOGITS

eering

0.97

 levers

0.90

 freak

0.85

 exercised

0.84

eer

0.80

 exerted

0.79

eers

0.78

 valve

0.77

orship

0.76

 cooker

0.76

Activations Density 0.854%

keywords related to control and authority in various contexts

No Comments

No Known Activations